Seven-coordinate iron and manganese complexes with acyclic and rigid pentadentate chelates and their superoxide dismutase activity

Article abstract of DOI:10.1021/ic7012039

The reactions of seven-coordinate [Fe^III(dapsox)(H ₂O)₂]Cl0₄·H₂O (1), [Fe ^II(H₂dapsox)(H₂O)₂](NO ₃)₂·H₂O (2), and [Mn^II(H ₂dapsox)(CH₃OH)(H₂O)](ClO₄) ₂(H₂O) (3) complexes of the acyclic and rigid pentadentate H₂dapsox ligand [H₂dapsox = 2,6- diacetylpyridinebis(semioxamazide)] with superoxide have been studied spectrophotometrically, electrochemically, and by a submillisecond mixing UV/vis stopped-flow in dimethyl sulfoxide (DMSO). The same studies were performed on the seven-coordinate [Mn^II(Me₂[15]pyridinaneN ₅)(H₂O)₂]Cl₂·H₂O (4) complex with the flexible macrocyclic Me₂[15]pyridinaneN ₅ ligand (Me₂[15]pyridinaneN₅ = trans-2,13-dimethyl-3,6,9,12,18-pentaazabicyclo-[12.3.1]octadeca-1(18),14, 16-triene), which belongs to the class of proven superoxide dismutase (SOD) mimetics. The X-ray crystal structures of 2-4 were determined. All complexes possess pentagonal-bipyramidal geometry with the pentadentate ligand in the equatorial plane and solvent molecules in the axial positions. The stopped-flow experiments in DMSO (0.06% of water) reveal that all four metal complexes catalyze the fast disproportionation of superoxide under the applied experimental conditions, and the catalytic rate constants are found to be (3.7 ± 0.5) × 10⁶, (3.9 ± 0.5) × 10⁶, (1.2 ± 0.3) × 10⁷, and (5.3 ± 0.8) × 10⁶ M^-1 s^-1 for 1-4, respectively. The cytochrome c McCord-Fridovich (McCF) assay in an aqueous solution at pH = 7.8 resulted in the IC₅₀ values (and corresponding k_McCF constants) for 3 and 4, 0.013 ± 0.001 μM (1.9 ± 0.2 × 10⁸ M^-1 s^-1) and 0.024 ± 0.001 μM (1.1 ± 0.3 × 10⁸ M^-1 s^-1), respectively. IC₅₀ values from a nitroblue tetrazolium assay are found to be 6.45 ± 0.02 and 1.36 ± 0.03 μM for 1 and 4, respectively. The data have been compared with those obtained by direct stopped-flow measurements and discussed in terms of the side reactions that occur under the conditions of indirect assays.

Full text of DOI:10.1021/ic7012039

Inorg. Chem. 2007, 46, 8825−8835
Seven-Coordinate Iron and Manganese Complexes with Acyclic and
Rigid Pentadentate Chelates and Their Superoxide Dismutase Activity
Gao-Feng Liu,† Milosˇ Filipovic´,‡ Frank W. Heinemann,† and Ivana Ivanovic´-Burmazovic´*,†
Institute for Inorganic Chemistry, UniVersity of Erlangen-Nu¨rnberg, Egerlandstrasse 1,
91058 Erlangen, Germany, and Department of Chemistry, UniVersity of Belgrade,
11000 Belgrade, Serbia
Received June 20, 2007
The reactions of seven-coordinate [FeIII(dapsox)(H2O)2]ClO4
[MnII(H2dapsox)(CH3OH)(H2O)](ClO4)2(H2O) (3) complexes of the acyclic and rigid pentadentate H2dapsox ligand
[H2dapsox 2,6-diacetylpyridinebis(semioxamazide)] with superoxide have been studied spectrophotometrically,
electrochemically, and by a submillisecond mixing UV/vis stopped-flow in dimethyl sulfoxide (DMSO). The same
studies were performed on the seven-coordinate [MnII(Me2[15]pyridinaneN5)(H2O)2]Cl2
H2O (4) complex with the
flexible macrocyclic Me2[15]pyridinaneN5 ligand (Me2[15]pyridinaneN5 trans-2,13-dimethyl-3,6,9,12,18-pentaazabicyclo-
[12.3.1]octadeca-1(18),14,16-triene), which belongs to the class of proven superoxide dismutase (SOD) mimetics.
The X-ray crystal structures of 2 4 were determined. All complexes possess pentagonal-bipyramidal geometry
‚ ‚H2O (2), and
H2O (1), [FeII(H2dapsox)(H2O)2](NO3)2
)
‚
)
−
with the pentadentate ligand in the equatorial plane and solvent molecules in the axial positions. The stopped-flow
experiments in DMSO (0.06% of water) reveal that all four metal complexes catalyze the fast disproportionation of
superoxide under the applied experimental conditions, and the catalytic rate constants are found to be (3.7
×
±
0.5)
4, respectively. The cytochrome
) 7.8 resulted in the IC50 values (and corresponding
1
1
106, (3.9
± 0.5) × ± 0.3) × ± 0.8) × −
106, (1.2 107, and (5.3 106 M- s- for 1
c McCord Fridovich (McCF) assay in an aqueous solution at pH
−
1
kMcCF constants) for 3 and 4, 0.013
±
0.001
µ
M (1.9
±
0.2
×
108 M- s-1) and 0.024
±
0.001
µ
M (1.1
±
0.3
×
±
1
108 M- s-1), respectively. IC50 values from a nitroblue tetrazolium assay are found to be 6.45
± 0.02 and 1.36
0.03
µM for 1 and 4, respectively. The data have been compared with those obtained by direct stopped-flow
measurements and discussed in terms of the side reactions that occur under the conditions of indirect assays.
Introduction
cause aging, pain, inflammatory disorders, serious neurode-
generative diseases, and multiple types of cancer.1b,2b,3
Therefore, the concept of removing superoxide via rapid
disproportionation, i.e., dismutation (eq 1), has a protective
beneficial outcome in a large number of diseases caused by
the overproduction of superoxide radicals.3a,4
Superoxide (O2•-) is the reactive radical anion formed
following a one-electron reduction of dioxygen during
numerous oxidation reactions under normal conditions in
both living and nonliving systems.1 Because it is a very good
reducing agent in the anionic form and a very good oxidant
in the protonated form [pKa(HO2) ) 4.69], superoxide is
potentially dangerous for all cellular macromolecules and
can generate other undesired reactive species.1,2 Its damaging
effects lead to different pathophysiological conditions that
2O2•- + 2H+ f H2O2 + O2
(1)
Natural superoxide dismutase (SOD) enzymes catalyze
reaction (1) and in preclinical and clinical trials have shown
promising therapeutic properties, although they suffer as drug
* To whom correspondence should be addressed. E-mail: ivana.
ivanovic@chemie.uni-erlangen.de.
† University of Erlangen-Nu¨rnberg.
(3) (a) Muscoli, C.; Cuzzocrea, S.; Riley, D. P.; Zweier, J. L.; Thiemer-
mann, C.; Wang, Z.-Q.; Salvemini, D. Br. J. Pharmacol. 2003, 140,
445-460 and references cited therein. (b) Stadtman, E. R. Curr. Med.
Chem. 2004, 11, 1105-1112.
(4) (a) Salvemini, D.; Riley, D. P.; Cuzzocrea, S. Nat. ReV. Drug DiscoVery
2002, 1, 367-374. (b) Cuzzocrea, S.; Thiemermann, C.; Salvemini,
D. Curr. Med. Chem. 2004, 11, 1147-1162.
‡ University of Belgrade.
(1) (a) Fridovich, I. J. Biol. Chem. 1989, 264, 7761-7764. (b) Fridovich,
I. Ann. N.Y. Acad. Sci. 1999, 893, 13-28.
(2) (a) Liochev, S. I.; Fridovich, I. IUBMB Life 1999, 48, 157-161. (b)
Salvemini, D.; Muscoli, C.; Riley, D. P.; Cuzzocrea, S. Pulm.
Pharmacol. Ther. 2002, 15, 439-447 and references cited therein.
10.1021/ic7012039 CCC: $37.00
Published on Web 09/20/2007
© 2007 American Chemical Society
Inorganic Chemistry, Vol. 46, No. 21, 2007 8825

Liu et al.
that of the native mitochondrial MnSOD.6a,8g At the same
time, these complexes are the first enzyme mimetics tested
in humans.
Scheme 1
In the case of the macrocyclic MnII mimetics (Scheme 1),
it has been postulated that the profound conformational
rearrangements of the macrocyclic pentadentates facilitate
subsequent electron transfer and that the ligands with high
conformational flexibility can assist SOD activity.8a,g Seven-
coordinate FeIIISOD mimetics with the same macrocyclic
chelate systems show a different catalytic mechanism in
which the aqua-hydroxo form of the complex, [Fe(L)(OH)-
(H2O)]2+, is the catalytically active species.11 A drawback
of these complexes is the low-pKa values of the two
coordinated water molecules, which results in the formation
of inactive (inert) dihydroxo complexes at the physiological
pH.11 Therefore, the idea is to design a chelate that will
decrease the acidity of the iron center and so increase the
concentration of the catalytically active aqua-hydroxo
species at the physiological pH to promote an enhanced SOD
activity. Because free iron ions are more toxic than manga-
nese ions,5 it is important that the chelate will form a very
stable complex and prevent the release of iron ions. Despite
this toxicity, complexes of FeIII would be highly attractive
as SOD mimetics because of their higher kinetic and
thermodynamic stability than MnII complexes.
Because we have shown that the conformational flexibility
of the pentadentate ligand is not a key requirement for the
SOD activity of the seven-coordinate complexes, due to the
fact that in an interchange substitution mechanism (operating
in the case of these complexes)12a efficient formation of a
real six-coordinate (with pseudo octahedral geometry) in-
termediate is generally not required, we were interested in
additional experimental validation of such a mechanistic
paradigm. This has been achieved by probing the reactivity
of appropriate conformationally inflexible complexes toward
superoxide.
candidates primarily from immunogenic response.2a,3a,4a,5 This
calls for new types of free-radical inhibiting enzyme mimetics
to be used as pharmaceuticals. Stable low molecular weight
metal complexes that can react with superoxide and ef-
ficiently replicate the activity of the native SOD enzyme have
the potential to become a new generation of drugs for the
treatment of diseases of various aetiologies.2b,3a,4a,5,6
Among the many different complexes that have been
studied as potential SOD mimetics,2b,3a,5,6b,7 the most efficient
synthetic SOD catalysts known to date are seven-coordinate
complexes of MnII with macrocyclic pentadentate chelates
derived from carbon-substituted pentaazacyclopentadecane
[15]aneN5 (Scheme 1).2b,3a,4a,5,6a,8 Their catalytic rate constants
were obtained by direct kinetic measurements, as the only
reliable method for quantitative assessment of activity,9,10
showing that the SOD activity of these complexes can exceed
(5) Riley, D. P. Chem. ReV. 1999, 99, 2573-2587 and references cited
therein.
(6) (a) Salvemini, D.; Wang, Z.-Q.; Zweier, J. L.; Samouilov, A.;
Macarthur, H.; Misko, T. P.; Currie, M. G.; Cuzzocrea, S.; Sikorski,
J. A.; Riley, D. P. Science 1999, 286, 304-306. (b) Vujaskovic, Z.;
Batinic-Haberle, I.; Rabbani, Z. N.; Feng, Q.-F.; Kang, S. K.;
Spasojevic, I.; Samulski, T. V.; Fridovich, I.; Dewhirst, M. W.;
Anscher, M. S. Free Radical Biol. Med. 2002, 33, 857-863 and
references cited therein. (c) Samlowski, W. E.; Petersen, R.; Cuzzocrea,
S.; Macarthur, H.; Burton, D.; McGregor, J. R.; Salvemini, D. Nat.
Med. 2003, 9, 750-755. (d) Okado-Matsumoto, A.; Batinic-Haberle,
I.; Fridovich, I. Free Radical Biol. Med. 2004, 37, 401-410.
(7) (a) Batinic-Haberle, I.; Spasojevic, I.; Hambright, P.; Benov, L.;
Crumbliss, A. L.; Fridovich, I. Inorg. Chem. 1999, 38, 4011-4022.
(b) Ohtsu, H.; Shimazaki, Y.; Odani, A.; Yamauchi, O.; Mori, W.;
Itoh, S.; Fukuzumi, S. J. Am. Chem. Soc. 2000, 122, 5733-5741. (c)
Li, D.; Li, S.; Yang, D.; Yu, J.; Huang, J.; Li, Y.; Tang, W. Inorg.
Chem. 2003, 42, 6071-6080. (d) Durackova, Z.; Labuda, J. J. Inorg.
Biochem. 1995, 58, 297-303. (e) Liao, Z.; Xiang, D.; Li, D.; Mei,
F.; Yun, F. Synth. React. Inorg. Met.-Org. Chem. 1998, 28, 1327-
1341. (f) Spasojevic´, I.; Batinic´-Haberle, I.; Stevens, R. D.; Hambright,
P.; Thorpe, A. N.; Grodkowski, J.; Neta, P.; Fridovich, I. Inorg. Chem.
2001, 40, 726-739. (g) Yamaguchi, S.; Kumagai, A.; Funahashi, Y.;
Jitsukawa, K.; Masuda, H. Inorg. Chem. 2003, 42, 7698-7700. (h)
Batinic´-Haberle, I.; Spasojevic´, I.; Stevens, R.; Hambright, D.; Neta,
P.; Okado-Matsumoto, A.; Fridovich, I. J. Chem. Soc., Dalton Trans.
2004, 1696-1702. (i) Shearer, J.; Long, L. M. Inorg. Chem. 2006,
45, 2358-2360. (j) Durot, S.; Lambert, F.; Renault, J.-P.; Policar, C.
Eur. J. Inorg. Chem. 2005, 2789-2793.
In this paper, we have synthesized and characterized seven-
coordinate FeII (2) and MnII (3) complexes of acyclic and
rigid pentadentate H2dapsox [H2dapsox ) 2,6-diacetylpyri-
dinebis(semioxamazide)]12b that have several important
features regarding their potential SOD activity. The reactivity
of these two complexes and the previously reported FeIII
complex (1) of the same ligand12c-e (all of which have the
structure shown in Scheme 2) toward superoxide has been
studied spectrophotometrically, electrochemically, and by a
submillisecond mixing UV/vis stopped-flow in DMSO. The
(10) Weiss, R. H.; Flickinger, A. G.; Rivers, W. J.; Hardy, M. M.; Aston,
K. W.; Ryanll, U. S.; Riley, D. P. J. Biol. Chem. 1993, 268 (31),
23049.
(11) Zhang, D.; Busch, D. H.; Lennon, P. L.; Weiss, R. H.; Neumann, W.
L.; Riley, D. P. Inorg. Chem. 1998, 37, 956-963 and references cited
therein.
(12) (a) Dees, A.; Zahl, A.; Puchta, R.; van Eikema Hommes, N. J. R.;
Heinemann, F. W.; Ivanovic-Burmazovic, I. Inorg. Chem. 2007, 46,
2459-2470. (b) Andjelkovic, K.; Ivanovic, I.; Prelesnik, B. V.; Leovac,
V. M.; Poleti, D. Polyhedron 1996, 15 (24), 4361-4366. (c)
Andjelkovic, K.; Bacchi, A.; Pelizzi, G.; Jeremic, D.; Ivanovic-
Burmazovic, I. J. Coord. Chem. 2002, 55, 1385-1392. (d) Ivanovic-
Burmazovic, I.; Hamza, M. S. A.; van Eldik, R. Inorg. Chem. 2002,
41, 5150-5161. (e) Ivanovic-Burmazovic, I.; Hamza, M. S. A.; van
Eldik, R. Inorg. Chem. 2006, 45, 1575-1584.
(8) (a) Riley, D. P.; Weiss, R. H. J. Am. Chem. Soc. 1994, 116, 387-
388. (b) Riley, D. P.; Henke, S. L.; Lennon, P. J.; Weiss, R. H.;
Neumann, W. L.; Rivers, W. J.; Aston, K. W.; Sample, K. R.; Rahman,
H.; Ling, C.-S.; Shieh, J.-J.; Busch, D. H.; Szulbinski, W. Inorg. Chem.
1996, 35, 5213-5231. (c) Riley, D. P.; Lennon, P. J.; Neumann, W.
L.; Weiss, R. H. J. Am. Chem. Soc. 1997, 119, 6522-6528. (d) Riley,
D. P.; Henke, S. L.; Lennon, P. J.; Aston, K. Inorg. Chem. 1999, 38,
1908-1917. (e) Riley, D. P. AdV. Supramol. Chem. 2000, 6, 217-
244. (f) Kra¨mer, R. Angew. Chem. 2000, 112, 4641-4642; Angew.
Chem., Int. Ed. 2000, 39, 4469-4470. (g) Aston, K.; Rath, N.; Naik,
A.; Slomczynska, U.; Schall, O. F.; Riley, D. P. Inorg. Chem. 2001,
40, 1779-1789.
(9) Riley, D. P.; Rivers, W. J.; Weiss, R. H. Anal. Biochem. 1991, 196,
344-349.
8826 Inorganic Chemistry, Vol. 46, No. 21, 2007

SeWen-Coordinate Iron and Manganese Complexes
Scheme 2
Bio-Kine software version 4.23 and also the Specfit/32 program.
At least 10 kinetic runs were recorded under all conditions, and
the reported rate constants represent the mean values.
Cyclic voltammetry measurements have been carried out using
an Autolab instrument with a PGSTAT 30 potentiostat. A conven-
tional three-electrode arrangement was employed consisting of a
gold disk working electrode (geometric area: 0.07 cm2) (Metrohm),
a platinum wire auxiliary electrode (Metrohm), and the Ag(s)/AgCl-
(s) wire as pseudo reference electrode, for the measurements in
DMSO, or a Ag/AgCl, NaCl (3 M) (Metrohm) reference electrode,
for the measurements in the aqueous solution [the potential vs NHE
was calibrated by the Ag/AgCl, NaCl (3 M), potential (0.222 vs
NHE)]. The measurements in DMSO were performed in the
presence of 0.1 M tetrabutylammonium hexafluorophosphate as the
supporting electrolyte, whereas the measurements in aqueous
solutions were done by applying a 0.1 M NaClO4 supporting
electrolyte. All solutions without superoxide were thoroughly
degassed with nitrogen prior to beginning the experiments, and
during the measurements, the nitrogen atmosphere was maintained.
Measurements with superoxide were carried out by saturating the
solution with dry oxygen ([O2] ) 2.1 mM).15 The sample
concentration was 0.5 mM. All experiments were performed at room
temperature.
catalytic SOD activity of the complexes with the acyclic and
rigid H2dapsox ligand has been compared (under the selected
experimental conditions) with the reactivity of a MnII (4)
complex with the flexible pyridine derivative of the [15]-
aneN5 macrocycle (Me2[15]pyridinaneN5 ) trans-2,13-
dimethyl-3,6,9,12,18-pentaazabicyclo[12.3.1]octadeca-1(18),-
14,16-triene), which belongs to the class of proven SOD
mimetics.8e The SOD activity of the complexes has also been
investigated in an aqueous solution by applying indirect
cytochrome c and nitroblue tetrazolium (NBT) assays, and
the data have been compared with those obtained by direct
stopped-flow measurements and discussed in terms of the
side reactions that occur under the conditions of indirect
assays.
Note! It should be pointed out that for DMSO solutions only
glass equipment and Hamilton Teflon ValVes can be used!
Safety Notes! Perchlorate salts of metal complexes with organic
ligands are potentially explosiVe. Only a small amount of material
should be prepared and handled with great care.
Synthesis of [FeII(H2dapsox)(H2O)2](NO3)2‚H2O (2). 2,6-
Diacetylpyridine (0.163 g, 1 mmol) and semioxamazide (0.218 g,
2.1 mmol) were mixed in 60 mL of a methanol/acetonitrile mixture
(2:1) and warmed up to 65 °C. The reaction mixture was refluxed
for 2 h under an argon atmosphere. Fe(NO3)3‚4H2O (0.313 g, 1
mmol) was carefully added to the resulting white suspension, and
its color changed to gray. After 1 h of refluxing, water (10 mL)
was added, resulting in a clear gray solution, which was cooled
down to room temperature and left standing for 5 h. The nice block
crystals were filtered off, washed with a small amount of acetone,
and dried in air (yield: 0.395 g, 71%). IR data (KBr, cm-1): 3509s
(NH), 3386s (H2O), 3129s, 2929m, 1715s, 1677s, 1614s, 1541m
(amide CdO), 1382s (amide), 1156m (CH), 822m, 681m. Anal.
Calcd for C13H21N9O13Fe: C, 27.53; H, 3.73; N, 22.22. Found: C,
27.74; H, 3.12; N, 22.49.
Synthesis of [MnII(H2dapsox)(CH3OH)(H2O)](ClO4)2(H2O)
(3). 2,6-Diacetylpyridine (0.163 g, 1 mmol) and semioxamazide
(0.218 g, 2.1 mmol) were added to 40 mL of methanol, and the
mixture was stirred at 55 °C for 1 h. Mn(ClO4)2‚4H2O (0.326 g, 1
mmol) was added into the resulting white suspension. The solution
color changed to yellow, while some of the white powder still
remained undissolved. The addition of 50 mL of CH3CN resulted
in a clear yellow solution. Four days later, light-yellow crystals
were collected (yield: 0.393 g, 60%). IR data (KBr, cm-1): 3501s
(NH), 3361s (H2O), 3247s, 1720m, 1677s, 1529m (amide CdO),
1388s (amide), 1337s, 1147m (CH) 1048w, 819m, 737m, 669m.
Anal. Calcd for C14H23N7O15Cl2Mn: C, 25.69; H, 3.54; N, 14.99.
Found: C, 25.47; H, 3.24; N 14.71.
Experimental Section
Materials. All solid chemicals were of p.a. grade and used as
received without any further purification. [Zn(H2dapsox)(H2O)2]-
Cl2 used in the electrochemical measurements was synthesized
according to the published procedure.13 High-performance liquid
chromatography (HPLC)-grade dimethyl sulfoxide (DMSO) con-
taining a controlled amount of water (0.06% after mixing in a
stopped-flow cuvette) was used for the complex solutions, and the
water content was determined by Karl-Fischer titration. KO2
solutions were prepared according to published procedures.14
Instrumentation and Measurements. Carlo Erba elemental
analyzers 1106 and 1108 were used for chemical analysis. IR and
UV/vis spectra were recorded on a Mattson FTIR 60 AR (KBr
pellets) spectrophotometer and a Hewlett-Packard 8542A spectro-
photometer, respectively.
Time-resolved UV/vis spectra were recorded on a modified Bio-
Logic stopped-flow module µSFM-20 (10 ms dead time) combined
with a Huber CC90 thermostat and equipped with a J&M TIDAS
high-speed diode array spectrometer with a combined deuterium
and tungsten lamp (200-1015 nm wavelength range). Isolast
O-rings were used for all sealing purposes to enable measurements
in DMSO. The spectrum of DMSO was used as a reference for all
spectroscopic measurements. For the rapid kinetic measurements,
the Bio-Logic stopped-flow module was upgraded to a submilli-
second mixing stopped-flow configuration by combining it with a
microcuvette accessory (with an optical path light of 0.8 mm) and
a monochromator to minimize the dead time of the instrument.
Measurements with the FeII complex were performed under an
atmosphere of dry nitrogen. Data were analyzed using the integrated
Synthesis of [MnII(Me2[15]pyridinaneN5)(H2O)2]Cl2‚H2O (4).
3,6-Diazaoctane-1,8-diamine (1.46 g, 1.0 mmol) was added drop-
wise to a hot solution of MnCl2‚2H2O (1.61 g, 1.0 mmol) and
2,6-diacetylpyridine (1.63 g, 1.0 mmol) in water (50 mL). After 4
(13) Sumar, M.; Ivanovic-Burmazovic, I.; Hodzic, I.; Andjelkovic, K. Synth.
React. Inorg. Met.-Org. Chem. 2002, 32, 721-737.
(14) Duerr, K.; Macpherson, B. P.; Warratz, R.; Hampel, F.; Tuczek, F.;
Helmreich, M.; Jux, N.; Ivanovic-Burmazovic, I. J. Am. Chem. Soc.
2007, 129, 4217-4228.
(15) Kryatov, S. V.; Rybak-Akimova, E. V.; Schindler, S. Chem. ReV. 2005,
105, 2175-2226.
Inorganic Chemistry, Vol. 46, No. 21, 2007 8827

Liu et al.
Table 1. Crystal Data and Structure Refinement for 2-4
2
3
4
empirical formula
fw
C13H21N9O13Fe
567.24
C14H21Cl2N7O14Mn
637.22
C15H33Cl2N5O3Mn
457.30
cryst size (mm3)
T (K)
0.25 × 0.16 × 0.10
0.18 × 0.09 × 0.08
0.42 × 0.38 × 0.35
100(2)
100(2)
295(2)
cryst syst
space group
a (Å)
b (Å)
c (Å)
monoclinic
P21/c
monoclinic
P21/c
orthorhombic
Pbca
10.245(1)
19.031(1)
22.756(1)
90
8.0028(6)
14.6046(5)
18.226(2)
93.079(9)
2127.1(3)
4
14.632(1)
11.257(1)
14.908(2)
99.148(6)
2424.3(4)
4
â (deg)
V (Å3)
4436.8(5)
8
Z
F(000)
1168
1.771
0.799
54411
1300
1.746
0.848
54150
1928
1.369
0.859
9042
F
calcd (Mg/m3)
µ (mm-1
)
reflns colld
indep reflns
5491
5346
4521
obsd reflns [I g 2σ(I)]
data/restrains/param
GOF
4700
5491/4/388
1.034
4300
5346/17/443
1.059
2403
4521/ 0/318
0.986
R1 [I g 2σ(I)]
wR2 (all data)
0.0262
0.0671
0.0357
0.0845
0.0520
0.1189
2
2
2
a R1 ) ∑||Fo| - |Fc||/∑|Fo|; wR2 ) [∑w(Fo - Fc )2/∑w(Fo )2]1/2
.
h of refluxing, the reaction mixture was filtered and the deep-orange
solution was allowed to cool down to room temperature. Deep-
orange crystals of the compound [MnIIL(H2O)2]Cl2 (L ) 2,13-
dimethyl-3,6,9,12,18-pentaazabicyclo[12.3.1]octadeca-1(18),2,12,-
14,16-pentaene) were obtained and dried under vacuum at 70 °C
for 3 h (2.29 g, 69% yield). This complex (2.2 g, ca. 5 mmol) was
dissolved in 40 mL of anhydrous EtOH, and the flask was
flushed with argon for a few minutes. NaBH4 (20 mmol, ca. 4
equiv/double bond) was added to the orange solution in one portion,
and the suspension was stirred at room temperature under argon.
Two hours later, the temperature was increased to 60 °C and the
mixture was stirred for the next 3 h. After cooling to room
temperature, the solvent was removed from the pale-yellow mixture.
The residue was dissolved in water (20 mL), and NaCl (4.5 g) was
added. The aqueous solution was extracted with CH2Cl2 (3 × 40
mL). The combined organic phases were dried (MgSO4) and
filtered, and the solvent was removed. The pale-yellow solid was
dissolved into 5 mL of water. Colorless crystals suitable for X-ray
structure analysis were obtained after 2 days (yield: 0.86 g, 18%
calculated from 2,6-diacetylpyridine). IR data (KBr, cm-1): 3524s
(NH), 3460s, 3300s (H2O), 3210s, 2911m, 2868s, 1628m, 1597m,
1579m (amide CdO), 1458s, 1381m (amide), 1126m, 1106s,
1009m, 1106s (CH), 1009m, 980s, 808m, 680m. Anal. Calcd for
C15H33Cl2MnN5O3: C, 39.40; H, 7.27; N, 15.31. Found: C, 39.38;
H, 7.31; N, 15.29.
X-ray Crystal Structure Determinations. Data for 2 and 3 were
collected at 100 K using a Bruker-Nonius Kappa CCD diffracto-
meter (λ ) 0.710 73 Å; graphite monochromator), while data for
4 were collected at room temperature using a Siemens P4 four-
circle diffractometer (λ ) 0.710 73 Å; graphite monochromator).
All data sets were corrected for Lorentz and polarization effects.
Absorption effects were taken into account by semiempirical
methods using either multiple scans (SADABS)16a for 2 and 3 or
the ψ-scan technique16b for 4. The structures were solved by direct
methods and refined using full-matrix least-squares procedures on
F2 (SHELXTL NT 6.12).16c The perchlorate anion in 3 is disordered;
two alternative positions have been refined, resulting in occupancies
of 52.0(6)% for O11-O14 and 48.0(6)% for O11′-O14′. With the
exception of the hydrogen atoms of the two methyl groups in 4,
which are in calculated positions of optimized geometry, the
positions of all other hydrogen atoms in 2-4 were derived from
difference Fourier maps. The isotropic displacement parameters of
all hydrogen atoms were tied to those of the equivalent isotropic
displacement parameters of their corresponding carbon, nitrogen,
or oxygen carrier atoms. Crystal data, data collection parameters,
and refinement details of the structure determinations of complexes
2-4 are summarized in Table 1; selected bond distances and angles
are listed in Table 2.
Indirect SOD Assays. (a) Cytochrome c Assay. SOD activities
of complexes were measured using standard McCord-Fridovich
assay17 based on a ferricytochrome c reduction with superoxide
produced by xanthine/xanthine oxidase. The assay was performed
at 25 °C in 3 mL of a reaction buffer (50 mM potassium phosphate
buffer; pH ) 7.8) containing ferricytochrome c (10 µM), xanthine
(100 µM), and an amount of xanthine oxidase such as to give a
rate of ∆OD550 nm ≈ 0.02 min-1 (about 0.01 U mL-1) in the absence
of a putative SOD mimic. A reduction of ferricytochrome c was
monitored at 550 nm. After 150 s, different amounts of the putative
SOD mimic were added. Rates were linear for at least 8 min. Both
rates in the absence and in the presence of the complex were
determined for each concentration of complex added and plotted
against it. The IC50 value represents the concentration of putative
SOD mimic that induces a 50% inhibition of the reduction of
cytochrome c.
(b) Reliability of the McCord-Fridovich Assay. To check that
the tested compounds do not inhibit the production of superoxide
by xanthine oxidase, the rate of conversion of xanthine to urate
(see below) was determined by measuring the increase in the
absorbance at 290 nm over a 2 min period with and without the
tested compounds. To measure the rate of conversion of xanthine
to urate, xanthine oxidase (20 µL of 1 U mL-1 XO) was added to
a solution of 50 mM potassium phosphate buffer, pH ) 7.8,
containing xanthine (150 µM) at a final volume of 1.0 mL at
25 °C. Urate production was monitored at 290 nm.18 No difference
(16) (a) SADABS, version 2.06; Bruker AXS, Inc.: Madison, WI, 2002.
(b) North, A. C. T.; Phillips, D. C.; Mathews, F. S. Acta Crystallogr.
1968, A24, 351-359. (c) SHELXTL NT 6.12; Bruker AXS, Inc.:
Madison, WI, 2002.
(17) McCord, J. M.; Fridovich, I. J. Biol. Chem. 1969, 244, 6049-6055.
8828 Inorganic Chemistry, Vol. 46, No. 21, 2007

SeWen-Coordinate Iron and Manganese Complexes
Table 2. Selected Bond Distances (Å) and Angles (deg) for 2-4
Compound 2
Fe1-O5
Fe1-O1
Fe1-N1
Fe1-N5
O3-C12
N6-C12
N2-N3
N5-N6
2.144(1)
2.179(1)
2.206(2)
2.218(2)
1.232(2)
1.343(2)
1.375(2)
1.372(2)
Fe1-O6
Fe1-O3
Fe1-N2
2.159(1)
2.195(1)
2.233(2)
O1-C8
N3-C8
N2-C6
N5-C10
1.234(2)
1.346(2)
1.289(2)
1.287(2)
Ã5-Fe1-O6
Ã5-Fe1-O1
O6-Fe1-O1
Ã5-Fe1-O3
O6-Fe1-O3
O1-Fe1-O3
Ã5-Fe1-N1
O6-Fe1-N1
O1-Fe1-N1
O3-Fe1-N1
N5-Fe1-N2
178.88(4)
88.10(4)
93.01(4)
91.79(4)
88.37(4)
77.12(3)
90.03(4)
89.16(4)
140.93(4)
141.95(4)
139.95(4)
Ã5-Fe1-N5
O6-Fe1-N5
O1-Fe1-N5
O3-Fe1-N5
N1-Fe1-N5
Ã5-Fe1-N2
O6-Fe1-N2
O1-Fe1-N2
O3-Fe1-N2
N1-Fe1-N2
91.96(4)
87.04(4)
148.91(4)
71.80(4)
70.15(4)
88.53(4)
91.92(4)
71.14(4)
148.23(4)
69.80(4)
Figure 1. ORTEP view of [FeII(H2dapta)(H2O)2(NO3)]+ in the crystal of
2 drawn with thermal ellipsoids at 50% probability level. Another nitride
anion is omitted for clarity.
Compound 3
Mn1-Ã5
Mn1-O6
Mn1-O1
Mn1-N5
O1-C8
2.152(2)
2.192(2)
2.224(4)
2.281(2)
1.242(3)
1.283(3)
1.379(2)
1.338(3)
Mn1-N1
Mn1-N2
Mn1-O3
2.287(2)
2.306(2)
2.359(2)
of 1 mg of solid KO2 was added into 2 mL of 50 mM potassium
phosphate buffer, pH ) 7.8, containing putative SOD mimetic, and
after 2 min, spectra were recorded. NBT reacted with superoxide,
forming blue pigment formazan [λmax ≈ 580 nm (35 000 M-1
cm-1)].19b The presence of the complex caused concentration-
dependent inhibition of formazan formation, as followed by the
absorbance change at 580 nm. The concentration that causes 50%
of the formazan formation was indicated as IC50.
O3-C12
N5-C10
N5-N6
1.227(3)
1.281(3)
1.371(2)
1.347(3)
N2-C6
N2-N3
N3-C8
N6-C12
Ã5-Mn1-O6
Ã5-Mn1-O1
O6-Mn1-O1
Ã5-Mn1-N5
O6-Mn1-N5
O1-Mn1-N5
Ã5-Mn1-N1
O6-Mn1-N1
O1-Mn1-N1
N5-Mn1-O3
171.94(7)
91.62(6)
90.87(6)
87.48(7)
86.89(6)
153.70(6)
93.00(7)
90.25(7)
138.17(6)
69.40(6)
N5-Mn1-N1
Ã5-Mn1-N2
O6-Mn1-N2
O1-Mn1-N2
N5-Mn1-N2
N1-Mn1-N2
Ã5-Mn1-O3
O6-Mn1-O3
O1-Mn1-O3
68.09(6)
95.90(7)
92.16(6)
69.78(6)
136.46(6)
68.39(6)
87.57(6)
85.05(6)
84.31(5)
Results and Discussion
Structures. The cationic [FeII(H2dapsox)(H2O)2]2+ com-
plex (2; Figure 1) has a pentagonal-bipyramidal (PBP)
structure with a neutral pentadentate H2dapsox ligand in the
equatorial plane and two water molecules in axial positions.
H2dapsox is coordinated to the FeII center through the
pyridine nitrogen, two imine nitrogen atoms, and two oxygen
atoms of hydrazide CdO groups. These five donor atoms
are almost perfectly coplanar, and the mean deviation from
planarity is just 0.0137 Å. The central FeII ion is only 0.0034-
(5) Å below this plane. The sum of four chelate angles and
the bite O1-Fe1-O3 angle (Table 2) is 360.01°, very close
to 360° for an ideal planar structure. Two axial water
molecules (O5 and O6) complete PBP, forming an almost
linear angle [O5-Fe1-O6 ) 178.88(4)°]. The intraligand
bond lengths (Table 2) suggest that the neutral H2dapsox
ligand is coordinated to the FeII center in a hydrazide >Cd
NsNHsCdO form, different from the R-oxiazine >
CdNsNdCsO- form of the deprotonated dapsox2- ligand
present in the PBP structure of the corresponding FeIII ([FeIII-
(dapsox)(H2O)2]+) complex.12c In the FeII complex, the
average C-N distance is a bit longer (1.316 Å), whereas
the average C-O (1.233 Å) and N-N (1.374 Å) distances
are shorter than those found in the FeIII complex (1.296,
1.280, and 1.392 Å, respectively). Despite the change in the
metal-ion size and ligand charge, there is just a small increase
Compound 4
Mn1-O2
Mn1-N1
Mn1-O1
Mn1-N4
2.241(3)
2.278(3)
2.282(3)
2.320(3)
Mn1-N3
Mn1-N2
Mn1-N5
2.330(4)
2.343(3)
2.352(3)
O2-Mn1-N1
O2-Mn1-O1
N1-Mn1-O1
O2-Mn1-N4
N1-Mn1-N4
O1-Mn1-N4
O2-Mn1-N3
N1-Mn1-N3
O1-Mn1-N3
N4-Mn1-N3
N2-Mn1-N5
89.0(2)
173.4(2)
85.3(2)
83.7(2)
142.4(2)
102.9(2)
99.1(2)
143.3(2)
83.6(2)
74.3(2)
140.5(2)
O2-Mn1-N2
N1-Mn1-N2
O1-Mn1-N2
N4-Mn1-N2
N3-Mn1-N2
O2-Mn1-N5
N1-Mn1-N5
O1-Mn1-N5
N4-Mn1-N5
N3-Mn1-N5
86.0(2)
70.2(2)
89.0(2)
145.2(2)
74.7(2)
95.0(2)
70.3(2)
86.2(2)
73.7(2)
143.2(2)
in the slope was recorded with or without the putative SOD mimics.
To exclude the possibility of hydrogen peroxide interference with
the assay10 and with the intent to avoid catalase addition (which
can make the system more complex), catalase activity of complexes
was monitored as described previously.18 No catalase-type activity
of our complexes was detected.
(c) Modified NBT Assay. To further probe SOD activity of our
complexes, a modified NBT assay was used.7i,19a In this essay, an
extensive excess of superoxide against catalyst was used. A total
(19) (a) Dutta, S.; Padhye, S.; Ahmed, F.; Sarkar, F. Inorg. Chim. Acta
2005, 358, 3617-3624. (b) Bielski, B. H. J.; Shiue, G. G.; Bajuk, S.
J. Phys. Chem. 1980, 84, 830.
(18) Policar, C.; Durot, S.; Lambert, F.; Cesario, M.; Ramiandrasoa, F.;
Morgenstern-Badarau, I. Eur. J. Inorg. Chem. 2001, 1807-1818.
Inorganic Chemistry, Vol. 46, No. 21, 2007 8829

Liu et al.
in the average Fe-N equatorial bond length from the FeIII
complex to the corresponding FeII complex (ca. 0.022 Å).
The same was observed in the case of the macrocyclic iron
PBP complexes, where the small effect was explained in
terms of the rigid nature of the cyclic ligands.20 However,
the acyclic nature of our ligand suggests that the weak
sensitivity of the Fe-N bonds to the change in the metal
oxidation state is a more general feature of the unsaturated
segments of the pentadentate ligands. From another side, the
average Fe-O equatorial bond length is significantly longer
in the FeII (2.187 Å) complex than in the FeIII (2.056 Å)
complex. The negative charge of the coordinated R-oxiazine
oxygen atoms additionally strengths the FeIII-O bond. The
prominent elongation of the average Fe-OH2 bond length
from the FeIII complex (2.028 Å) to the FeII complex (2.152
Å) is observed, confirming that the axial distances reflect
changes in the ion size more readily than the equatorial
ones.21 Interestingly, the asymmetry in the two Fe-N(imine)
bonds in [FeII(H2dapsox)(H2O)2]2+ (0.015 Å) is somewhat
more prominent than that in [FeIII(dapsox)(H2O)2]+ (0.007
Å). However, it is ca. 2 times smaller than that in the case
of the very similar [FeII(H2dapsc)(H2O)(Cl)]+ complex (0.034
Å), where the difference in these two bonds was rationalized
in terms of the high-spin d6 configuration and the Jahn-
Teller effect in a PBP field.21 The differences in the two
equatorial Fe-O bonds are identical (ca. 0.02 Å) in all three
structures ([FeIII(dapsox)(H2O)2]+, [FeII(H2dapsox)(H2O)2]2+,
and [FeII(H2dapsc)(H2O)(Cl)]+) and are not affected by the
change in the iron oxidation state and charge of the oxygen
atom. The average Fe-N and Fe-O equatorial bonds in
[FeII(H2dapsox)(H2O)2]2+ and [FeII(H2dapsc)(H2O)(Cl)]+ are
almost identical. Even more, the average Fe-OH2 bond
length in [FeII(H2dapsox)(H2O)2]2+ and the corresponding
bond length in [FeII(H2dapsc)(H2O)(Cl)]+ are also identical.
This shows that the axial coordination of Cl- does not
affect the bonds either in the equatorial plane or in its trans
position.
Figure 2. ORTEP view of [MnII(H2dapsox)(H2O)(CH3OH)(ClO4)]+ in
the crystal of 3 drawn with thermal ellipsoids at 55% probability level.
Another perchloride anion is omitted for clarity.
ingly, this interaction between ClO4- and amide tails of the
tweezer-like cationic complex is even observed in the MeCN
solution of the FeIII as well as the FeIII form of the complex.22
The [MnII(H2dapsox)(CH3OH)(H2O)]2+ complex (3; Figure
2) also has the PBP structure with the neutral pentadentate
ligand coordinated in the equatorial plane in a hydrazide
>CdN-NH-CdO form. Thus, the intraligand bond lengths
are almost identical with those in the above-discussed FeII
complex (Table 2). However, the Mn-N and Mn-O bonds
are all longer than the corresponding Fe-N and Fe-O bonds
(Table 2). The sum of the chelate angles and the bite angle
is 359.97°, which means that the five donor atoms from H2-
dapsox form an ideal planar structure. In fact, the mean
deviation of this plane is just 0.0195 Å, with the central MnII
ion a little below this plane [the distance from the plane is
ca. 0.0412(8) Å].
In comparison with the similar structure of [MnII(H2dapsc)-
(H2O)(Cl)]+,21 it can be seen that, even though the average
Mn-N bond length is the same for both structures (ca. 2.291
Å), the average Mn-O equatorial bond length is significantly
longer in our complex (2.291 vs 2.216 Å), showing that there
is no general interrelation between the equatorial distances,
as was suggested. Even more, the difference between two
Mn-O equatorial bond lengths is very significant (0.135 Å),
although such a distortion is not expected for the spherically
symmetrical high-spin d5 electronic configuration. This shows
that the intra- and intermolecular secondary interactions
within the crystal packing have an important influence on
the symmetry of the coordination sphere. As in the case of
2, the intramolecular hydrogen bonds are observed between
terminal -NH2 groups and the corresponding metal-
coordinated hydrazide oxygen atoms [N4-H‚‚‚O1 2.703(2)
Å and N7-H‚‚‚O3 2.763(2) Å] and between the hydrazide
nitrogen and amide oxygen atoms [N3-H‚‚‚O2 2.749(2) Å
and N6-H‚‚‚O4 2.630(2) Å]. As mentioned above, an
The two terminal -NH2 groups are involved in intramo-
lecular hydrogen bonds with the metal-coordinated hydrazide
oxygen atoms [N4-H‚‚‚O1 2.666(2) Å and N7-H‚‚‚O3
2.714(2) Å; given are only the donor-acceptor distances for
the hydrogen bridges discussed, and complete details can
be found in Table S1 in the Supporting Information], similar
to those observed in the structure of [FeIII(dapsox)(H2O)2]+.12c
The presence of the hydrogen atom on the hydrazide nitrogen
enables the formation of additional intramolecular hydrogen
bonds between the hydrazide nitrogen atoms N6 or N3 and
the amide oxygen atoms O2 and O4 [N3-H‚‚‚O2 2.793(2)
Å and N6-H‚‚‚O4 2.692(2) Å]. Although hydrogen-bond
-
interaction between the cationic complex and the NO3
counteranion is observed [N7-H‚‚‚O11 3.002(2) Å and N4-
H‚‚‚O11 3.486(2) Å], the ion-pair association in this structure
is not as prominent as it is in the case of the interactions
-
-
between ClO4 and [FeIII(dapsox)(H2O)2]+, where ClO4
participates in an extensive hydrogen-bond network. Interest-
(20) Drew, M. G. B.; Hamid bin, O. A.; Nelson, S. M. J. Chem. Soc.,
Dalton Trans. 1976, 1394.
(21) Palenik, G. J.; Wester, D. W. Inorg. Chem. 1978, 17, 864-870.
(22) Sarauli, D.; Popova, V.; Zahl, A.; Puchta, R.; Ivanovic´-Burmazovic´,
I. Inorg. Chem. 2007, in press.
8830 Inorganic Chemistry, Vol. 46, No. 21, 2007

SeWen-Coordinate Iron and Manganese Complexes
Figure 3. ORTEP view of [MnII(Me2[15]pyridinaneN5)(H2O)2]2+ in the
crystal of 4 drawn with thermal ellipsoids at 25% probability level. The
chloride anion is omitted for clarity.
Figure 4. Cyclic voltammogramms of (a) 1 purged with nitrogen, (b) 1
purged with dioxygen, (c) 1 purged with dioxygen, and (d) pure DMSO
purged with dioxygen. Conditions: [complex] ) 0.5 × 10-3 M; [Bu4NBF4]
) 0.1 M; T ) 298 K; scan rates ) 0.2 V s-1
.
Scheme 3
Figure 5. Cyclic voltammogramm of 3 purged with dioxygen. Condi-
tions: [3] ) 0.5 × 10-3 M; [Bu4NBF4] ) 0.1 M; T ) 298 K, in a DMSO
interesting feature of the tweezer-like complexes of H2dapsox
is their association with the anion by means of shape
recognition, hydrogen-bond complementarity, and charge
assistance. The perchlorate anion is chelated by the cationic
[MnII(H2dapsox)(CH3OH)(H2O)]2+ complex via hydrogen
bonds [N4-H‚‚‚O23 2.924(3) Å, N7-H‚‚‚O24 3.136(3) Å,
and N7-H‚‚‚O22 3.578(3) Å], closing the cavity of the
complex and forming sort of a macrocyclic structure.
In order to compare the reactivity of our complexes with
that of a proven SOD catalyst under the selected experimental
conditions, we have synthesized and characterized the [MnII-
(Me2[15]pyridinaneN5)(H2O)2]2+ complex (4) with the flex-
ible pyridine derivative of the [15]aneN5 macrocycle, which
belongs to the class of proven SOD mimetics.8 Similar to
other [Mn([15]aneN5)]-type complexes,8e,g [MnII(Me2[15]-
pyridinaneN5)(H2O)2]2+ exists in the seven-coordinate PBP
geometry but crystallizes as a trans-diaqua complex instead
of a trans-dichloro complex. The Mn-N bond distances and
angles (Figure 3) are all quite similar to those observed for
other complexes of the same class. 4 crystallizes as a mixture
of (S,S)- and (R,R)-dimethyl enantiomers with a C-CH3 and
N-H pattern on the macrocycle, which alternates as up-
down-up-down. Consequently, the two sides of the mac-
rocyclic plane and the two axial coordination sites as well
are chemically equivalent. By way of comparison with the
solution; scan rates ) 0.2 V s-1
.
analogous imine groups containing complex,23 the average
Mn-N distance in 4 is somewhat longer because the double-
bond-containing ligand has a smaller cavity. The structure
of the FeIII complex with the same ligand is known,24 but
the ligand is present in its (R,S)-dimethyl diastereomer form
where two methyl groups are on the same side of the FeN5
plane.
Electrochemistry. The metal-centered redox potential is
the most important criterion for the complex to be a SOD
•-
mimetic because the catalytic disproportionation of O2
requires redox reactions between the complex and superoxide
(Scheme 3). The complex redox potential should fall between
•-
the redox potentials for the reduction and oxidation of O2
viz. -0.16 and +0.89 V vs NHE, respectively.25
,
Aqueous solutions of [Fe(dapsox)(H2O)2]ClO4 (1) in the
pH range of 1-12 exhibit a reversible redox wave for the
FeIII/FeII couple, and no complex decomposition or dimer-
ization was observed.26 Furthermore, in the pH range of
1-10, the metal-centered redox potential for [Fe(dapsox)-
(23) Omar, J.-S.; Daniel, R.-R.; del Mar´ıa, J. R.-H.; Martha, E. S.-T.; Rafael,
Z.-U. J. Chem. Soc., Dalton Trans. 1998, 1551-1556.
(24) Drew, M. G. B.; Rice, D. A.; Silong, S. B. Polyhedron 1983, 2, 1053-
1056.
Inorganic Chemistry, Vol. 46, No. 21, 2007 8831

Liu et al.
Figure 6. Time-resolved UV/vis spectra recorded for the reaction of a 5 × 10-5 M complex with 1 mM KO2 in DMSO at room temperature for 1 (a), 2
(b), 3 (c), and 4 (d). A: spectrum recorded (measurements in a tandem cuvette) before mixing. B: first spectrum obtained after mixing (using a stopped-flow
module) followed by spectra recorded at time intervals of 10 s (total observation time 2.5 h). Inset: control reaction without the addition of the complex
followed over 2.5 h.
(H2O)2]ClO4 is in the range required for the possible SOD
activity. For the FeIII complexes with [15]aneN5 types of
chelates that are proven SOD catalysts, the redox potentials
were not reported for the physiological pH, at which these
complexes exist as an equilibrium mixture of the dihydroxo
and aqua-hydroxo species.11 At pH ∼ 3, they have redox
potentials in the range of 0.35-0.45 V vs NHE. In
comparison, our complex at pH ) 3 and 7.8 shows reversible
redox behavior at 0.34 and 0.05 V vs NHE, respectively. It
should be noted that dapsox2- causes an increase in the pKa
values of the coordinated water molecules (pKa1 ) 5.8 and
pKa2 ) 9.5),12d,26 which are very close to those of the native
FeIIISOD enzyme (∼5 and ∼9).27 Thus, at the physiological
pH, almost 100% of the complex is in the catalytically active
aqua-hydroxo form.
FeII couple at slightly more negative potential, -0.18 V,
because in the presence of oxygen, it is more difficult to
reduce FeIII. When the scan proceeds toward more negative
potentials (Figure 4c), after the complex is reduced to the
FeII species, molecular oxygen is reduced to superoxide at
-0.82 V. When the scan is then returned to 0.2 V, no
•-
corresponding anodic peak assigned to the oxidation of O2
is found, in contrast to the reversible redox behavior for
dioxygen in DMSO solutions (Figure 4d). The intensity of
the anodic peak corresponding to the oxidation of FeII is also
significantly decreased. This suggests that the iron complex
(starting from the electrochemically generated FeII form)
catalytically decomposes superoxide (the FeII form, present
in lower concentration than O2•-, consumes all of it). The
superoxide decomposition is also observed by applying much
lower (catalytic) concentrations of the complex (Figure S1
in the Supporting Information).
Similar to the proven seven-coordinate MnIISOD mimetics
with [15]aneN5 types of chelates,8b 3 is stable in the pH range
of 6-10.5 and in methanol exhibits a reversible redox
potential at 0.8 V vs NHE.8a,b The redox behavior in aqueous
solutions for the macrocyclic MnSOD mimetics was not
reported. We measured the cyclic voltammograms for 4 (Eox
) 0.98 V and Ered ) 0.35 V) and 3 (Eox ) 0.64 V and Ered
) 0.20 V) at pH ) 7.8, and both complexes show similar
The cyclic voltammogram for 1 in DMSO purged with
nitrogen exhibits a reversible couple at -0.13 V vs Ag/AgCl
electrode (Figure 4a), or -0.11 vs NHE, obtained by
calibration with ferrocene. The cyclic voltammogram in
dioxygen-saturated DMSO in the scan range up to -0.4 V
(Figure 4b) shows again a reversible redox wave for the FeIII/
(25) Stanbury, D. M. AdV. Inorg. Chem. 1989, 33, 70-138.
(26) Sarauli, D.; Meier, R.; Liu, G.-F.; Ivanovic-Burmazovic, I.; van Eldik,
R. Inorg. Chem. 2005, 44, 7624-7633.
(27) Lah, M. S.; Dixon, M. M.; Pattridge, K. A.; Stallings, W. C.; Fee, J.
A.; Ludwig, M. L. Biochemistry 1995, 34, 1646-1660.
8832 Inorganic Chemistry, Vol. 46, No. 21, 2007

SeWen-Coordinate Iron and Manganese Complexes
Figure 7. Plots of kobs vs [complex] for the reaction between complexes and saturated KO2 in a DMSO solution at room temperature. Inset: kinetic trace
for [1] ) 2 × 10-4 M and control reaction without the addition of the complex obtained by using submillisecond mixing stopped-flow configuration.
behavior with large peak separation (Figure S2 in the
Supporting Information).
spectra (Figure 6) show that immediately after mixing of a
superoxide solution with a complex solution, rapid decom-
position of O2 (decrease in the absorbance in the 240-
•-
Electrochemical measurements in dioxygen-saturated DMSO
(Figure 5) show, as in the case of 1, that 3 can also
catalytically decompose superoxide (disappearance of the
anodic peak assigned to the oxidation of O2•- in the presence
of the MnII form of the complex), remaining unchanged
(appearance of a reversible redox wave for the MnII/MnIII
couple at 0.65 V). In comparison, the proven SOD mimetic
4 upon reaction with superoxide undergoes modification, and
in the scan range from 0 to 1.2 V and back to 0 V, three
oxidation and reduction peaks appear (Figure S3 in the
Supporting Information).
Control electrochemical measurements of the [Zn(H2-
dapsox)(H2O)2]2+ complex in nitrogen- and dioxygen-
saturated DMSO confirmed that H2dapsox was not the redox-
active ligand (the zinc complex is electrochemically silent)
and that in the presence of its non-redox-active metal
complex electrochemically generated superoxide was stable
(the anodic peak assigned to the oxidation of O2•- does not
disappear in the presence of the zinc complex).
330 nm range within the dead time of the stopped-flow
instrument) was observed. The products of superoxide
disproportionation, O2 and H2O2, were qualitatively detected
in all four experiments.28 In the case of 4, following fast
superoxide decomposition, the complex starts to decompose
slowly and results in the formation of a light-brown colloid
precipitate (presumably MnO2) after ∼3 h. Three hours after
mixing with KO2, acid was added to the solutions of
complexes 1-3, which resulted in the recovery of the initial
1 and 3 complexes, respectively. This demonstrates that our
acyclic complexes are more stable than the macrocyclic
complex 4 under the applied experimental conditions, which
is in agreement with the electrochemical observations (see
above).
The rapid process was quantified by following the corre-
sponding absorbance decrease at 270 nm in a series of
stopped-flow measurements, in which the catalytic concen-
tration of the studied complexes was varied. Application of
a microcuvette accessory (which reduced the dead time of
the instrument down to 0.4 ms) enabled observation of the
fast disappearance of the 270 nm absorption, which could
best be fitted as a first-order process to obtain the charac-
teristic kobs (s-1) value. When experiments were performed
using the complex solutions with a higher amount of water,
Reaction with Superoxide in DMSO. We studied the
•-
reactions of 1-4 (Figure 6), with a large excess of O2 in
DMSO containing a controlled amount of water (0.06%),
which was in excess over the superoxide and complex
concentrations. Water present in the DMSO solution plays
an important role and enables the catalytic decomposition
of O2•-. Similar to what was reported in the literature,8a under
absolute water-free conditions only a stoichiometric reaction
(28) For the O2 detection, see: Karlin, K. D.; Cruse, R. W.; Gultneh, Y.;
Farooq, A.; Hayes, J. C.; Zubieta, J. J. Am. Chem. Soc. 1987, 109,
2668-2679. For the H2O2 detection, a peroxide indicator paper suitable
for the organic solvents (QUANTOFIX-Peroxide 100) was used.
•-
between O2 and the complex could be observed and the
catalytic process was suppressed. Time-resolved UV/vis
Inorganic Chemistry, Vol. 46, No. 21, 2007 8833

Liu et al.
Table 3. Catalytic Rate Constants and IC50 Values Obtained by Using
Direct Stopped-Flow Measurements in DMSO (0.06% Water) and an
Indirect Cytochrome c Assay in an Aqueous Solution (Phosphate Buffer
pH ) 7.8), Respectively
complex
k
cat. (M-1 s-1
)
IC50 (µM)
kMcCF (M-1 s-1
)
MnSOD
1
2
0.005 ( 0.001
5.2 ( 0.2 × 108
(3.7 ( 0.5) × 106
(3.9 ( 0.5) × 106
(1.2 ( 0.3) × 107
(5.3 ( 0.8) × 106
3
0.013 ( 0.001
0.024 ( 0.001
1.9 ( 0.2 × 108
1.1 ( 0.3 × 108
4a
a kcat. ) 1.0 × 107 M-1 s-1 was obtained in an stopped-flow experiment
with DMSO/H2O ) 1/18.8a-d,9
it was not possible to quantify the corresponding rate
constants because the higher water contents caused a mixing
problem on a short time scale. In Figure 7, the obtained kobs
values are reported as a function of the complex concentra-
tion for our iron and manganese complexes, as well as for
the proven SOD mimetic. A good linear correlation between
kobs and the complex concentration was observed for all
studied complexes. From the slope of the plot of kobs vs
9,10
catalyst concentration, the catalytic rate constants (kcat.
)
were determined (Table 3). The kcat. values show that both
iron complexes and the macrocyclic manganese complex
have almost the same catalytic activity, within the error
limits, whereas 3 has approximately 2 times higher activity.
It should be stressed that it does not matter whether the FeIII
or FeII form of the complex is used; identical spectral changes
(Figure 6) and kinetic behavior (Figure 7) are observed upon
reaction with superoxide, which is consistent with the redox
cycling of the complex during O2•- decomposition (Scheme
3).
Reaction with Superoxide in Aqueous Solutions. To
prove the reactivity of our complexes toward superoxide in
an aqueous buffer, a McCord-Fridovich assay (here referred
to as a X/XO assay)17 and modified NBT7i,19a assays were
used.
Figure 8. (a) Kinetics of the reduction of ferricytochrome c (550 nm)
without and with the putative SOD mimics. Reduction of ferricytochrome
c (followed during 300 s) caused by the addition of the indicated
concentrations of tested complexes and MnSOD (E. coli). (b) Inhibition
percentage as a function of the concentrations of 4 and 5: IC50 determi-
nation.
We also have observed that the catalytic rate constants
obtained by using a X/XO assay (Table 3) are at least 1 order
of magnitude higher than those obtained by a stopped-flow
method. We found this to be due to the direct reaction
between complexes and cytochrome c. When higher complex
concentrations than those that cause nearly 100% inhibition
were used, a reoxidation of reduced cytochrome c was
observed, suggesting that the oxidized form of the complex
(generated during the catalytic cycle) acts as an oxidant of
cytochrome c.31
The X/XO assay is based on kinetic competition for the
superoxide reaction between oxidized cytochrome c and the
complex showing SOD activity. The reduction of ferricyto-
chrome c was followed spectrophotometrically at 550 nm.
Both 3 and 4 were found to inhibit the reduction of
cytochrome c when injected into the solution, as shown in
Figure 8a. Inhibition percentages were measured for several
complex concentrations (Figure 8b), and IC50 values, cal-
culated using the graphical method, are reported in Table.
3. Although used as a feature characterizing the SOD activity
of the complex, IC50 values strongly depend on the concen-
tration of the detector used and are thus not appropriate for
comparison with the literature.18 From the measured IC50
values, it is possible to calculate a catalytic constant (kMcCF),
which is independent of the detector concentration. At the
IC50 concentration, superoxide reacts at the same rate with
the detector and with the putative SOD mimic. Then, kMcCF
) kdetector[detector]/IC50.29 Upon a comparison of the obtained
The iron complex exhibits an opposite effect, acting in its
reduced form as a reductant of cytochrome c, and increases
(30) (a) Liao, Z.-R.; Zheng, X.-F.; Luo, B.-S.; Shen, L.-R.; Li, D.-F.; Liu,
H.-L.; Zhao, W. Polyhedron 2001, 20, 2813-2821. (b) Rodriguez,
M.-C.; Morgenstern-Badarau, I.; Cesario, M.; Guilhem, J.; Keita, B.;
Nadjo, L. Inorg. Chem. 1996, 35, 7804-7810. (c) Gauuan, P. J. F.;
Trova, M. P.; Gregor-Boros, L.; Bocckino, S. B.; Crapo, J. D.; Day,
B. J. Bioorg. Med. Chem. 2002, 3013-3021. (d) Kitajima, N.; Osawa,
M.; Tamura, N.; Morooka, Y.; Hirano, T.; Hirobe, M.; Nagano, T.
Inorg. Chem. 1993, 32, 1879-1880. (e) Xiang, D. F.; Tan, X. S.;
Hang, Q. W.; Tang, W. X.; Wu, B.-M.; Mak, T. C. W. Inorg. Chim.
Acta 1998, 277, 21-25. (f) Spasojevic, I.; Batinic-Haberle, I.; Stevens,
R. D.; Hambright, P.; Thorpe, A. N.; Grodkowski, J.; Neta, P.;
Fridovich, I. Inorg. Chem. 2001, 40, 726-739. (g) Kachadourian, R.;
Batinic-Haberle, I.; Fridovich, I. Inorg. Chem. 1999, 38, 391-396.
(31) Detailed kinetic investigations are in progress.
kMcCF constants with those found in the literature,6,18,30
3
belongs to the very active SOD mimetics.
(29) Butler, J.; Koppenol, W. H.; Margoliash, E. J. Biol. Chem. 1982, 257,
10747-10750.
8834 Inorganic Chemistry, Vol. 46, No. 21, 2007

SeWen-Coordinate Iron and Manganese Complexes
the slope of cytochrome c reduction when injected into
solution (Figure 8a).31 Thus, the calculation of its IC50 and
kMcCF were not possible. To prove its activity in aqueous
medium, we used a modified NBT assay. Here, instead of
xhanthine/xhantine oxidase system, KO2 was used as source
of superoxide. When 1 was present in solution, an inhibition
of blue formazan formation was observed in a concentration
dependent manner. The concentration of 1 that caused 50%
inhibition of formazan formation (followed at 580 nm) was
6.45 µM. 4 caused 50% inhibition at the concentration of
1.36 µM, while 3 showed no effect. However, we observed
that 3 reacts with NBT itself, presumably forming a new
complex that, if the higher complex concentrations (>1 ×
10-3 M) were used, immediately precipitated as a yellow
powder. The solution of 1 (>1 × 10-3 M) gets slightly milky
in the presence of NBT after 1 day, whereas no interaction
between 4 and NBT was observed. Interactions with the NBT
indicator account for no detectable SOD activity of 3 and
somewhat lower activity of 1 than would be expected based
on the stopped-flow measurements.
ligand, is not the rate-limiting step in the overall inner-sphere
catalytic SOD pathway of the proven macrocyclic SOD
mimetics.12a Furthermore, it also shows that conformational
flexibility of the pentadentate ligand is not the key factor
assisting SOD activity and that the acyclic and rigid ligand
systems can also be considered as structural motifs for the
design of SOD mimetics. An additional advantage can be
the fact that their syntheses are more economic than the
syntheses of macrocyclic ligands.
We have also shown that the indirect SOD assays, which
are the most used methods for demonstrating complex SOD
activity, are not very reliable10 and can be applied only upon
considering possible cross-reactions between the indicator
substance and the studied complex in their different oxidation
forms, in which they may occur within the SOD catalytic
cycle. The direct stopped-flow method, where a high excess
of superoxide over the complex can be utilized, is a better
probe for a complex SOD activity even though it requires a
DMSO medium. Importantly, as was stressed by Sawyer et
al., an even closer relation between the kinetic measurements
in aprotic media than in bulk water can be drawn with the
processes in mitochondria, which are the major source of
superoxide in the aerobic organisms, because aprotic media
“may be representative of a hydrophobic biological matrix”.33
Under less protic conditions, causing a longer half-life of
O2•-, efficient superoxide decomposition is even more
desirable.
Although neither of the indirect methods we used proved
to be reliable, they show, in the manner utilized in the
literature,6,18,30 that our complexes exhibit substantial SOD
activity in aqueous solutions as well.
Conclusions
Although it has been postulated in the literature that only
seven-coordinate complexes of macrocyclic ligands with
prominent conformational flexibility could possess SOD
activity,8g,32 our seven-coordinate iron and manganese com-
plexes with the acyclic and rigid H2dapsox ligand demon-
strate the ability for catalytic decomposition of superoxide.
Similar to what usually was found in the case of the
macrocyclic pentadentate ligands,11 the manganese complex
shows higher SOD activity than the corresponding iron
complex. However, higher stability of the iron complex over
a very wide pH range is an advantage in terms of a possible
application.
Acknowledgment. The authors gratefully acknowledge
financial support from the Deutsche Forschungsgemeinschaft
within SFB 583 “Redox-active Metal Complexes”. We are
also thankful to Prof. R. van Eldik for helpful discussions.
Supporting Information Available: X-ray crystallographic files
in CIF format for the structure determinations of complexes 2-4,
a table listing the analysis of hydrogen bonds for the crystal
structures of 2 and 3, and cyclic voltammogramms for 1 (variable
concentrations) in DMSO purged with dioxygen, for 3 and 4 in
an aqueous solution, and for 4 in DMSO purged with dioxy-
gen. This material is available free of charge via the Internet at
http://pubs.acs.org.
The demonstrated SOD activity of the rigid seven-
coordinate complexes is in agreement with our recent
findings that water release and formation of a six-coordinate
intermediate, requiring conformational rearrangement of the
IC7012039
(33) Chin, D.-H.; Chiericato, G., Jr.; Nanni, E. J., Jr.; Sawyer, D. T. J.
Am. Chem. Soc. 1982, 104, 1296-1299.
(32) Riley, D. P. AdV. Inorg. Chem. 2007, 59, 233-263.
Inorganic Chemistry, Vol. 46, No. 21, 2007 8835